Image-Caption Model Based on Fusion Feature

Round 1

Reviewer 1 Report

This article describes a new image caption model. Though the topic have many real world applications, I found following weaknesses.

1) There are many captitalization (Upper case) errors. For example, the starting letter of some words are capitalized in the middle of a sentence.  For example, in the second paragraph of Introduction there exist "...... task was to realize modal conversion from Image mode to text mode through the model, which was embodied as binary group (I, S), and the model completed modal conversion from  Image mode I (Image) ...". Here, I in Image mode needs to be in lower case. Severe number of such type of errors exist in the literature. Recheck the whole manuscript for such type of errors. In the second sentence of abstract, capitalize the 'C' in the convolutional neural network. Check the whole manuscript. Only capitalize the first letter of the first word of a particular sentence. Only first letter of subject/nouns in the middle of a sentence needs to be capitalized. Otherwise, not required. There exist severe errors.

2) There exist large number of full stop errors. For example, again in the second paragraph of Introduction there exist "The task of image captioning is done by Farhadi[1] ,et al proposed that their task was to realize modal conversion from......". Here, you should write properly as "....Farhadi et al. [1] proposed that their task was.....". Alyas capitalize the first character of the author, then put a full stop after "al", then put a space, then write the citation. Check the whole manuscript for such type of similar errors. Unwanted full stops are placed in the middle of a sentence.

3) Always put a space before citing an article. i.e., put a space before [.]. Similarly put a space before first abbreviation i.e., before (.).

4) Many sentences are confusing and ambiguous. Rewrite the whole manuscript.

5) Include a paragraph at the end of the Introduction section as follows. "The remaining part of the article is structured as follows. Section 2 describes related work, whereas Section 3......"

6) Clearly describe the significance of BLEU-1, BLEU-4, CIDEr, METEOR, and ROUGE-L in a subsection. . If possible specify the difference among these metrics.

7) Some more recent methods (2021/2022) need to be compared with the proposed method. If possible include different type of comparisons between method.

8) Remove "6. Paten" that is present after Table 1. I think wrongly added.

9) Put the Table 2 caption at top of the Table not below the Table. Figure captions are placed at the bottom of the Figure whereas Table captions are place at the top of the Table.

10) Figure caption and figure should be in a single page. Correct this for Figure 5.

11) Add more images in Figure 5. Only 2 is very less.

12) Clearly highlight your contribution in Introduction.

13) Rewrite the whole manuscript for any grammatical,  full stop, comma, space, citation, upper-case and lower-case errors.

 

"

 

 

 

 

Author Response

Please see the attachment。

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript has some typos and alignment issues, so authors should check the paper completely and avoid jargon words.

Authors should describe more the importance of the Image Caption model and fusion feature process and how it will be connected with the current problems.

The literature review section is very poor, authors should add recent works done on the fusion feature models, squeeze and extraction, and appropriate machine learning techniques that can be used. Authors may use the references like Addressing Binary Classification over Class Imbalanced Clinical Datasets Using Computationally Intelligent Techniques

Enhancement of Detection of Diabetic Retinopathy Using Harris Hawks Optimization with Deep Learning Model

It is very much required to show the detailed process of the Data preprocessing using Min-Max normalization and also how the feature extraction process is done.

Authors should give a clear representation of the usage of equations 2 and 3 and its description is needed for a better understanding of the readers.

Authors must show why the equations from 10 to 15 are used in the decoder process and why the model should adopt a double-layer memory network.

Authors can elaborate on the results section with respect to the accuracy, quality of prediction, and precision.

The conclusion should be modified as per the results achieved in the proposed system and also authors should highlight by comparing the appropriate results with the existing systems.

Author Response

Please see the attachment。

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Though many of my concerns are incorporated, many grammatical mistakes are there. For example, in the last paragraph of the Introduction, it should be "The remaining parts of the article ...." not "The remaining part of the articles .....". Please check the manuscript in Grammarly, at least the basic grammatical errors will be identified.

Yes, I appreciate your efforts as this is your first english article. But, I request you to please check your paper in Grammarly at least once. 

Figure 1 is appeared twice. In page 4 without caption and in Figure 5 with caption. Delete the figure present in Fig 4. Check the whole manuscript for any errors.

Finally, take your time. No need to be in hurry, and please read the manuscript once, you will find the errors. After this, your paper will be more attractive.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

English language and style are fine/minor spell check required

Author Response

Please see the attachment.

Author Response File: Author Response.pdf